Protein secondary structure prediction via kernel minimum squared error

In this paper, we propose a new protein secondary structure prediction method based on kernel minimum square error (KMSE). KMSE is a supervised pattern classification method, which has been successfully applied to a wide range of pattern recognition problems. The naive KMSE focuses on two-class problem, so it can not be directly applied for protein secondary structure prediction. We design a multi-class classifier based on KMSE for protein secondary structure prediction. The results of our experiments carried out on the rs126 dataset show that the performance of our method is better than that of PCA and LDA. Our method achieves a very high degree of prediction accuracy with simple computation, and we believe it is an effective method for the prediction of the secondary structure of protein. © 2011 IEEE.